The present invention relates to Computer and communication networks and, more particularly, to methods and apparatuses for an intelligent scalable switching network.
The explosive growth of the Internet and the world wide web""s increasing multimedia content is placing great demand on the telco local loop infrastructure. The advent of Internet telephony coupled with the increasing need for video services, such as video conferencing, video mail, video on demand, and Television over IP, promises to grow into an unprecedented market opportunity An integrated telecommunications device capable of supporting Internet telephony, high speed Internet access, next generation video services, and data communications is urgently needed to expand broadband residential and remote services. The present invention substantially fulfills this need.
The present invention provides methods and apparatus for implementing an intelligent scalable switching network. In one aspect, the present invention is an apparatus comprising a unified Internet portal server (UIP Server) having multi-line capability, and a unified Internet portal client (UIP Client) incorporating functionality of a Customer Premise Equipment (CPE) DSL Modem, wherein the UIP Client is capable of communicating with the UIP Server via a network to provide a service to a subscriber using the UIP Client. According to the invention, the UIP server comprises a single server chassis incorporating all of a plurality of processing elements. In addition, the UIP Server is located remote from a Subscriber Location (SL) and is capable of providing a plurality of services using a Digital Subscriber Line (DSL). The UIP Client is located at the SL and is capable of deploying DSL capability.
In another aspect, the present invention provides a method for providing an intelligent scalable network. According to the invention, the method comprises (a) providing, at a location remote from a Subscriber Location (SL), a unified Internet portal server (UIP Server) having multi-line switching capability, wherein the UIP Server is capable of providing a plurality of services using a Digital Subscriber Line (DSL); (b) incorporating all of a plurality of processing elements within a single UIP Server chassis; (c) providing, at a Subscriber Location (SL), a unified Internet portal client (UIP Client), wherein the UIP Client is capable of deploying DSL capability; (d) incorporating functionality of a Customer Premise Equipment (CPE) DSL Modem in the UIP Client; and (e) providing for communication of the UIP Client to the UIP Server via a network, whereby the UIP Server is capable of providing a service to a subscriber using the UIP Client.
As discussed more fully below, use of the present invention allows for many features and advantages, depending on the configuration of the various embodiments of the present invention.
One of the innovative characteristics of an embodiment of the present invention is that a multi-line switching capability is incorporated. The incorporation of the multi-line switching capability gives this embodiment the capability of effecting a digital cross connect functionality within a single server chassis of the present invention. The digital cross connect functionality enables it to statistically multiplex the subscriber side interface in an efficient manner.
The digital cross connect capability relies on the principal that only a few subscribers need to actively be on the network at a given time, whereas a larger number of subscribers would at some time like to have connectivity to the network. The advantage of the over-subscription due to the digital cross connect is that the service provider is able to provide dial up connectivity to the network for a larger pool of subscribers, then he would be able to achieve with a dedicated connection methodology.
Embodiments of the present invention incorporate a multi-DSL configurational capability on a per line basis. A number of different types of DSL exist including but not limited to, SDSL which is Symmetric Digital Subscriber Line, ADSL which is Asymmetric Digital Subscriber Line, IDSL which is ISDN based Digital Subscriber Line. Each of these different DSL technologies have different characteristics, such as what bandwidth they can sustain, what reach they are able to accomplish DSL, and what line coding methodologies they use. For example, HDSL is a high bit rate DSL. VDSL is a higher speed based DSL. Each different DSL type addresses different requirements for different subscribers. One of the advantages of having a per line multi-DSL architecture is that the service provider is able to provide a wider selection of services to a broader customer base and thereby increase his own business viability.
Prior art methodology deploys multi-DSL capability within a chassis framework only. Within prior art there is a common chassis based connectivity and different line cards capable of supporting different DSL standards, such as SDSL, ADSL, VDSL, HDSL or IDSL, plug into the chassis on a line card basis. A given line card supports either a specific DSL standard or may support more than one DSL standard. The innovation within the current invention is that the Multi-DSL support is on a per line basis, thus unlike the prior art, a given subscribers line interface can be configured for any of the DSL standards under software configurational control.
In the prior art, a given line is configured for only one of the DSL standards. Therefore, prior art chasses steer the data from the various line cards to achieve a Multi-DSL capability on a chassis basis. One of the limitations of the prior art architectural design is that multiple types of linecards are required to deploy multiple services. Another limitation is that the multi-DSL capability is not a per line Multi-DSL in that a given subscriber line is specifically configured to the a single DSL flavor. Another limitation of the prior-art architectural design is that since a subscriber is either a SDSL, or an ADSL or another DSL standard so that re-provisioning to the subscribers DSL Standard requires service provider intervention.
The present invention deploys Multi-DSL capability on a per line basis. The present invention accomplishes the per line Multi-DSL by utilizing a more flexible front end that is simultaneously capable of addressing the needs of a number of DSL types. One specific embodiment of the current invention supports three specific DSL standards, namely, SDSL, ADSL, and IDSL. The current invention can then re-provision the line in a dynamic manner as required by the subscriber to any of the three DSL standards. The architecture of the present invention enables the subscriber to be able to deploy, for example, SDSL during the day, when perhaps its business needs require symmetric bandwidth, and, in the evening switch over to ADSL, when perhaps its requirements are asymmetric such as video on demand, broadcast TV type of applications. Another advantage of the per line basis multi-DSL capability of the present invention is that no service provider intervention is required for the dynamic provisioning and that this service capability can be re-provisioned by the subscriber based on his own changing needs and requirements.
The architecture of one embodiment of the present invention is designed to be a self strapping architecture. One embodiment of the current invention incorporates a storage element, either in the form of a hard drive, semiconductor memory, or any other similar storage device. The storage element can either reside at the service provider premises or can alternatively reside remotely from the service provider premises by use of a wide area storage area network. The storage element holds the configurational capabilities and the initialization parameters for the operation of the present invention. In case such a storage device is incapable of service to the present invention for a limited or extended period of time, the present invention incorporates a self strapping methodology so that a minimal functional capability within the digital switching network is configured and that the network is capable of providing continued operation for its subscribers.
The present invention incorporates a self strapping capability within the architecture which is capable of initialization of the system. The self strapping capability of the present invention can configure its own basic functionality, and is then able to configure its own network capability. Particularly and in one embodiment, if the network element is found at a later time, the system is then able to download parameters and requirements at that time and continue operation. The self strapping functionality within the present invention makes it possible for reliability of the system to be limited by the reliability of the self strapping capability of the present invention rather than the reliability of the storage element, or the reliability of the storage area network. In this manner, the present invention is capable of meeting stringent reliability requirements while simultaneously providing much higher functionality than otherwise possible due to the low mean time between failure (MTBF) of storage elements such as hard drives or other high density storage devices.
One embodiment of the present invention incorporates a storage element in the form of a local hard drive within the server chassis. The incorporation of the storage element provides a large storage device that is dedicated to a limited number of subscribers in a distributed manner. The innovative incorporation of the storage element makes it possible to provide an expanded service capability for the subscriber. The expanded capability enabled can be remote recording capability, video on demand capability, and other storage intensive capabilities.
The dedicated storage element incorporation differentiates this embodiment from the prior art in that a non-dedicated storage methodology such as a large network is not able to sustain the bandwidth and performance requirements of these storage intensive application. The storage element incorporation does not require a local hard drive but can rather be a network based storage area network. In the case of a storage area network the differentiating factor between prior art and the present invention is that there is a physical association of a storage element with a specific chassis. Moreover, the network topology is capable of supporting this relationship and performance so that the associated storage element of the present invention is able to get a guaranteed performance and delivery which is reasonably comparable to that of a local hard drive system. The present invention utilizes network based redundancy through out the network to achieve these goals.
Embodiments of the present invention incorporate the capability of interfacing with an ATM network. The incorporation of the ATM interface enables the server to interface with a large number of broadband data networks that exist within the public switched telephone network system infrastructure, such as broadband ISDN and the core switching capability within the PSTN. One embodiment of the present invention incorporates an ATM based interface such as OC3, OC12, or OC48 which enables the present invention to quickly and easily interface to the existing infrastructure. Within the scope of this document the ATM based interface OC3, OC12, OC48, OC96, and OC-N, where N is an integer, are used interchangeably. The use of ATM interfaces enables the present invention to deploy Quality of Service guarantees. Furthermore, ATM is a very efficient packet based architecture and is a flexible means of delivering widely varying bandwidth and traffic performance requirements.
Another embodiment incorporates the use of Multi Protocol Label Switching(MPLS) wherein a tag is associated with the VPI/VCI combination. The designated tag is then utilized to manage Quality of service guarantee and other traffic, routing, and performance related functions. Yet another embodiment of the present invention incorporates the use of Multi protocol over ATM(MPOA). MPOA in conjunction with MPLS enables support, traffic policing, bandwidth allocation, and general quality of service control for IP based traffic such as Internet, or other computer network.
One embodiment of the present invention incorporates a video farm element capability. In one such embodiment, the video farm element is a DVD based optical video player architecture where a carousal is used to provide access to a plurality of video disc based media. For example, an embodiment of the present invention utilizes a DVD based front end carousal which can select one of many DVD discs. The front end carousal after selecting the specific DVD can then insert it into one of a plurality of DVD Players. The Video Farm element of one embodiment of the current invention utilizes an ATM-based interface, such as OC3, OC12, or OC48, as well as a Digital Video Broadcast (DVB) cable based infrastructure to deliver the video streams resulting from the Video Farm Element. A single video farm element and multiple video farm elements can be interconnected in various topologies to enable access to a large video database on a on-demand basis.
Use of the optical video farm element is advantageous in that it is possible to configure the embodiment so that no hard drive storage is required to have a large variety of video stream delivery. The limitation of such a embodiment is that the video stream performance is limited. One embodiment of the present invention utilizes a combination of local hard drives within the video farm element, as well as a plurality of optical drives to achieve a balance between performance and cost.
The current invention also allows for incorporation of a standard cable TV interface. This interface enables efficient communication with existing broadcast infrastructures. In particular, a video farm element may be managed in a remote configuration. The output is then collated together in a DVB cable TV based head end. The head end can then transmit the video stream data using either satellite, hard wire or other means necessary to a host of other chassis residing either at the central office, a service provider premises, or at a subscriber premises. As provided herein, the terms Central Office (CO), service provider premises, central office site, and service provider UIP Server central location are used interchangeably.
The cable TV based video farm element broadcast stream as well as standard cable TV is incorporated to provide the capability of watching broadcast TV in an easy and seamless manner. Embodiments of the present invention utilize the video farm element broadcast stream a Cable TV based DVB-C interface, to concurrently deliver multi channel stream video. One embodiment of the present invention enables a specific chassis to tune to a number of digital sources and be able to isolate a single video stream on a per subscriber basis.
One embodiment of the present invention provides each subscriber his own digital tuner. In this embodiment, the subscriber is able to tune to any given broad cast signal within the network, thus the subscriber""s designated tuner is responsible for delivering the selected content. One such embodiment of the present invention provides a limited number of tuners that subscribers share on an over-subscription basis. Each tuner is able to tune to a plurality of digital streams concurrently, in this regard each subscriber is assigned his own sub-channel. In particular, the Digital Video Broadcast (DVB) Cable based technology utilized within this embodiment typically has a single MHz channel that is the fundamental tuned channel. This tuned analog channel carries a plurality of digital channels. One embodiment intelligently manages these tuned analog channels and assigns up to 8 subscribers within the tuned analog channel. Each subscriber is allocated his own digital channel. In this respect only three separate analog frequencies, or tuners, are required for a 24 subscriber system. This embodiment simplifies the tuning requirements at the subscriber premises and is a cost efficient way of getting full independent tuning capability. A disadvantage of this embodiment is that specific streams within a channel would be allocated to specific subscribers, and that there would be a fixed relationship between network elements and channels. In such a topology, it would be difficult to scale beyond a finite number of total subscribers. An embodiment of the present invention circumvents this scaling limitation by having segregated network segments so that a pool of n specific subscribers are able to tune to a given network segment. The present invention enables the service provider to use either a per subscriber based tuning topology or a per chassis based tuning topology to deploy his broadcast environment.
Certain embodiments of the present invention also incorporate an innovative means of multi-channel statistical multiplexing. Statistical multiplexing in these embodiments is used to enhance the performance of network based on the fact that not all users utilize the network to the same extent. The subscriber is dynamically allowed to take up as much bandwidth as he requires and then the whole pool of subscribers are then treated on a statistical basis thus each subscriber is able to get access to more capability than he would have access to on a dedicated basis. Temporary statistical aberrations occurring as a consequence of few subscribers over utilizing their resources can be absorbed by the current invention in this manner. In this manner, most subscribers get access to a much larger bandwidth when they utilize it then the network would be able to support on a continuous basis.
The present invention utilizes a multi-channel statistical multiplexing by being able to take a number of DSL channels and essentially treat them as a single pool of DSL channels as enabled by the architecture of the present invention. The resources are dedicated on a per chassis basis rather than being dedicated on a per channel basis. The advantage for the subscriber is that every user sees a pool of resources that seems much larger than he actually ends up utilizing statistically and therefore is able to realize a much better performance than on a dedicated per line basis.
This embodiment of the present invention is differentiated from prior art in that prior art does not utilize statistical multiplexing at the processing level. In particular, when a signal is received it needs to be modulated and demodulated. For certain DSL embodiments, such as ADSL, the processing portion of the modulation/demodulation represents a large portion of the total resource requirements of the channel. Prior art implements such functionality on a per line basis. The present invention shares this processing power as a pool of resources and the processing power is utilized in the most conservative manner. A large percentage of the processing capability, such as the Fourier transform and other processing intensive portions of wire signal modulation and demodulation are shared across a plurality of lines within the present invention.
The disadvantage of the prior art methodology in an always on DSL technology is that when the user is not utilizing his maximum bandwidth the processing capability is sitting idle. The present invention uses that residual idle processing capability to be able to generate a pool of processing capability and then allocates it on a required bandwidth basis. The present invention thru the use of multi-line statistical multiplexing and front end activity detectors provides the always on capability in addition to the shared processing pool and the resultant cost reduction.
The use of statistical multiplexing within the present invention enables the ability to provide dynamic provisioning capability so that every subscriber is able to provision his own services without the intervention of the service provider. In addition to the cost aspects, the present inventions innovative use of shared multi-line statistical multiplexing allows the subscriber to utilize the bandwidth as he needs it on a transport load bases rather than a maximum bandwidth basis. The subscriber is able to utilize the bandwidth as he needs it and at times he is able to peak and go over his general subscribed levels. The service providers deployment model is simplified in that he can simply monitor total processing utilization numbers and total load factors to determine over-subscription and resource allocation needs of the network.
The present invention by the use of multi-line statistical multiplexing therefore provides a new means and level of network performance prediction and control. The current invention enables a new tiered billing methodology based on a level of service agreement. For example, when the level of service agreement is exceeded, such as at peak time, a tiered billing agreement can exists and subscriber can be charged more based on exceeding the basic service agreement.
The present invention incorporates a means for enabling dynamic provisioning from the customer premises. In one embodiment of the present invention the subscriber is able to configure the performance of his services directly from the subscriber side to the service provider server. The present invention can achieve this dynamic provisioning capability without requiring any intervention from the service provider. The service provider configures the limiting parameters of the service offering, such as maximum bandwidth allowed, typical quality of service performance level, or the different type of DSL allowed. The subscriber then select either a semi dynamic profile or a dynamic profile. A semi dynamic profile enables the subscriber to simply select a given bandwidth and performance characteristic of his service and set this as the preferred configuration. A dynamic profile enables the subscriber to negotiate a specific bandwidth on a per application basis or on a per session basis.
The present invention is able to achieve the dynamic provisioning capability as a consequence of several innovations within the architecture. In one embodiment of the present invention, a multi-DSL capability is incorporated which enables the subscriber to choose the type of DSL selected. The subscriber can select the DSL flavor base on his immediate application requirements. The subscriber can select a symmetric DSL technology in case of symmetric applications thus if the subscriber wants to utilize a 1.5 megabit symmetric capability he is able to do so. Alternatively, if the user wants to select an asymmetric DSL technology in case of asymmetric application requirements he is able to do so.
The incorporation of the dynamic provisioning capability is advantageous to the service provider in that the network is able to defray congestion by not being mandated to maintain a service level when the subscriber is not utilizing any bandwidth. This provides a more expanded and unified service capability and enables the service provider to attract a larger customer base.
The use of dynamic provisioning, coupled with traffic conditions of the network and the time of day it is used, provide a tiered service rates structure that is capable of dynamically billing bandwidth usage. Within one embodiment of the present invention, symmetric bandwidth during the day costs more than symmetric bandwidth during the night. Thus, the service provider is able to maximize profits by aggressively targeting customer usage patterns.
One embodiment of the present invention also utilizes the dynamic provisioning capability to simplify deployment. The service provider configures all lines to be configured for a specific range of performance and DSL technology. When the subscribers"" line is installed and he subsequently establishes connectivity for the first time, the subscriber is presented with a dynamic provisioning browsing screen. The subscriber can select the specific features of his service at this time automatically. At this time special promotions can be offered to the subscriber by the service provider.
One embodiment of the present invention provides the capability of simultaneously supporting multiple services such as Video, Voice and Data within a single chassis. The ability of providing multi-service model is one of the enabling technologies for providing the unified communications model in the next section.
An embodiment of the present invention incorporates the capability for a unified communications methodology. The present invention enables the subscriber to innovatively access multiple services in an unified, easy and useful manner. The unified communications model also enables the capability of managing quality of service based on a applications levels. One embodiment of the present invention utilizes the unified communications model to specify quality of service classes based on the applications within the unified communication model and thereby manages quality of service across the network.
One embodiment of the present invention assigns a certain quality of service characteristics to the video on demand application, and different quality of service (QoS) characteristics to other services such as distance learning. Thus, although the same type of transport and network is used, differing QoS characteristics are easily managed. In one embodiment, the specific application within the unified communication model in conjunction with the subscribers selected performance levels are used to assign the QoS parameters. The association of the QoS with the application service is done at the subscriber since all services congregate at the subscriber and the subscriber is capable of managing performance characteristics dynamically thru use of dynamic provisioning.
Furthermore, an embodiment of the present invention has the capability to charge on a per service request basis. Depending on the subscriber""s service level agreement and his dynamic provisioning selection, one embodiment can charge the subscriber based on the application type used rather than the traffic type used.
Still further, one embodiment of the present invention incorporates the capability of a unified internet portal at the service provider side of the subscriber line interface. This embodiment is able to realize optimized performance from the subscriber perspective. In particular, when service data is available at one end of the wire and a request is made at the other end of the wire the transmission across the wire becomes the primary limiting factor, so long as devices at both ends of the wire are capable of supporting the application requirements adequately. Due to increasing computational speeds of personal computers the subscribers end is reasonably well equipped to adequately support the application demands. The embodiment described is specifically designed to support the necessary application requirements and is able to limit the network performance to simply the limitations of the wire interface.
The unified communication model makes it possible to access any service within the unified portal that are available to the subscriber in the highest performance manner possible within the network. Any data that is required beyond the unified portal is limited first by the same wire limitations but in addition such data is further limited by other network latencies and delays. The unified communication portal model always guarantees the highest performance possible for the subscriber. In addition to applications that require storage at the unified portal server, one embodiment utilizes the capacity of the unified portal server as a user cache. User data is stored within the network at the users server in a distributed manner for the group of subscribers. When the subscriber requests any specific services such as broadcast TV, video on demand, among others the content is delivered from the portal server to the subscriber directly.
In one embodiment, the unified portal is able to further the content down the wire right to the subscriber premise equipment. The limitation of pushing the data to the subscriber is that such configuration mandates an intelligent customer premise equipment, such as a computer with a hard drive, or a set top box. An intelligent customer premise equipment would increase the cost to the user over an unintelligent customer premise equipment. Furthermore, due to the dedicated nature of such a customer premise equipment and due to physical separation between individual customer premise equipment no sharing between multiple subscribers is generally possible. The service provider based unified portal residing at the service provider premise attracts a larger customer base because no high density storage is required at the subscriber premises, and thus inherently simplifies the deployment model for the subscriber. In one embodiment of the present invention a set top box has the capability of not only connecting to the network, but also of requesting the data either from the server or storing it within the customer premise equipment.
One embodiment of the present invention incorporates a switching and routing capability. In one form, the invention incorporates an ATM switch within the service provider server. In addition to the switching capability enabled by the use of the ATM switch, an embodiment of the present invention also incorporates a routing capability within the service provider server. The routing module can maintain either a localized routing table within the distributed architecture of the network or else comprehensive routing tables can be maintained within a centralized architecture. The local routing tables within the distributed service provider servers can be synchronized and maintained thru the use of the interconnection network. The distributed routing architecture utilizes a distributed routing table at the service provider server location to look up addresses that cannot be resolved thru the immediate history of the ATM switch.
In one embodiment, when an ATM connection is setup a VPI/VCI is assigned based on a routing lookup based structure. The VPI/VCI directs the incorporated switch and the switch at the next layer the information for forwarding the packet. The packet forwarding occurs at a reasonably high speed and thus the ATM switch has to be able to maintain the wire speed. The server architecture requires the performance rate of the ATM switch to be at least one or two orders of magnitude greater than the routing resolution performance.
The routing performance within one embodiment of the present invention is significantly slower than the packet switching capability. When the ATM switch is unable to resolve the packet path on a VPI/VCI basis, the whole packet is scheduled for a routing action. The server uses the VPI/VCI and the packet information to attempt to achieve a local routing resolution. The local routing resolution is a historical cache of previous routes resolved. If this embodiment is unable to resolve the packet path from the local tables, the system then attempts to resolve the packet path utilizing the comprehensive routing table, that may reside remotely. If the system is further unable to resolve the packet path, it then executes a sequence of algorithms to attempt to locate the receiving end of the packet. Upon successful resolution of the packet path, the system forwards this information to the ATM switch, the comprehensive routing table, and the local routing table to begin forwarding the packet and to store the forwarding information within its tables.
Another embodiment of the present invention incorporates a striping methodology for video on demand and general data striping. One embodiment of the present invention subdivides a stream of data into smaller pieces and distribute it in an intelligent manner across the network. The objective of the subdivision of data is that if the data is requested by multiple users at different times than a single server may be incapable of supporting the service load. By subdividing the data across the network the system is able to achieve a level of predictability and better network performance. Data striping is a means of load balancing the network so that if multiple subscribers serviced by a single server request an otherwise excessive service load then the service load is distributed across the network and no one server is overburdened.
This is accomplished in that the total load requirements are not solely the responsibility of the subscriber server, but rather are averaged across the network in a predetermined manner. One embodiment of the present invention attempts to optimize multiple levels of striping. In particular, striping occurs at the overall network level so that data is spread across a number of different servers across the network. As an example, if the network includes eight different servers and the required service happens to be a VOD movie than one eighth of each movie can-be striped across each of the servers. Thus, regardless of which specific subscriber requests a which specific movie, all servers will have an effective load of xe2x85x9th to be able to service any given request.
The striping within one such embodiment is done in such a manner that the first one eighth of the movie comes from the first server, the second eighth of the movie comes from the second server and so on. Consequently, there is a predictive behavior for the rest of the network in terms of what is going to occur at the next discrete time interval. For example, once a specific movie is requested, it is known ahead of time which server is responsible to deliver the first eighth of the movie. Furthermore, it is already known that server 2 for the specific movie will have a load after the time for the first eighth of the movie has elapsed.
In addition to the multi-server based striping methodology, one embodiment of the present invention expands on the striping model in that it includes the concept of a proxy server. When service of a striped data set is requested the system is able to make reliable prediction in regards to the subsequent loads from the various servers. Thus it is possible to project that a subsequent server will experience congestion based on the history of service requests. When congestion is anticipated the system can instead utilize a different proxy server instead of the assigned server and thereby circumvent the anticipated congestion. One embodiment of the present invention incorporates the proxy servers in a fashion so that different proxy servers are assigned to different data sets thus server n is not the proxy server for server 2 at all time, but rather any subset of proxy servers can be assigned as a proxy server for server 2 based on which data is requested by the subscriber. The striping capability in conjunction with the proxy server capability enables the present invention to balance the server load in a predictable and reliable manner.
In addition to the proxy server capability, one embodiment of the present invention also incorporates the capability of data striping within the specific storage devices within a single server. The same principle as multi-server striping is used but instead of using multiple servers the designated data is further subdivided into segments which are then striped across multiple storage elements within a given server. The upper limit to this methodology is the number of different storage devices that can reside within a single server. Assuming that a specific embodiment of the present invention has four hard drives, then the one eighth segment of the movie within the specific servers responsibility is further segmented into four sub-segments and each of the sub-segments reside within a different hard drive. The multi-storage data striping within one embodiment of the invention is a means of avoiding congestion at a given hard drive equipment level and achieving balanced data loading within a server. The multi-storage data striping can utilize different storage devices such as magnetic drives, optical drive, or semiconductor memory within a server. The use of semiconductor memory within the multi-storage striping additionally provides latency balancing in addition to load balancing. The first several seconds or minutes of a data segment are striped within semiconductor memory, such as dynamic ram or static ram. When a service request is received from the subscriber for a specific movie the initial few seconds are played from the semiconductor memory while the load tables are build and the network load balancing can occur. Any server within the network can be used to retain the first few seconds of the movie. In addition to the designated primary semiconductor memory server there also are proxy semiconductor memory servers distributed across the network that are responsible for alternatively providing the data striping few seconds within proxy caches. Consequently, this embodiment of the present invention is able to achieve very low latency between the time the subscriber requests a data service to the time the data service is delivered to the subscriber. While the short segment within semiconductor memory is exhausted, the system has computed the best scheduled bandwidth allocation based on the various load balancing algorithms and the sequence of server selection has been made to deliver a continuous data stream. In addition, in another embodiment of the present invention, the semiconductor memory and the hard drive can be alternatively used to store video clip, such as commercials and trailers, can be previews prior to the delivery of a continuous data stream.
One embodiment of the present invention incorporates a mechanism for load shedding within the network which further enhances the load balancing capability of the network. The present invention incorporates a intra-chassis load shedding mechanism by which a given subscriber can transparently be slotted to be managed by a different server within the network than its native server. The penalty of exercising this mechanism within one embodiment is that there is a slight latency penalty experience by the subscriber, however this is relatively insignificant in comparison to the poor bandwidth the subscriber would have experienced due to the congestion at its native server.
Tapered algorithms exist that are used to determine the level of load shedding so that as new subscribers come on line to the network, these new subscribers are then gradually negotiated thru the latent path to a non native server. In this manner, load shedding is achieved in a planned and simplified manner. The subscriber whose native server is experiencing a certain network conditions, will not be aware of any congestion, and will be completely unaffected.
One embodiment of the present invention implements load shedding at two different levels, at the line level and at the application level. In case of load shedding at the line level the processing of the line modulation and the modem capability is actually shed to a different box. Thus the native server is primarily only responsible for line interface functionality""s. In case of application level load shedding, such as voice over IP, the initial digitization of the line interface, the modulation and basic modem functionality is handled by the native server, however, the application level functionality is managed by a non-native server.
One embodiment of the current invention incorporates a lifeline service capability. Life line services are those services that are required to be operational when no power is available at the subscriber premises. Within prior art the life line services are carried only over the pots lines. In the pots only configuration the power required for the service is supplied from the central office to the subscriber. The subscriber can pickup the pots line, and be able to dial out. The amount of power utilized by the subscriber during the telephone call is minimal. One embodiment of the present invention innovatively extends the support of voice service during power outages to also support the multiple voice extensions. The multiple voice extensions within one embodiment utilize packet based technology. The present invention makes it possible to deliver packetized lifeline services at the subscriber site as well as the central office site. The central office life line capability is achieved by use of a 48 V DC battery backed supply as is standard within the industry. The innovation within the central office side of the server architecture is that only the voice support and basic DSL modem functionality is active during a power outage condition, and all the basic video and data services are intelligently place in a standby mode. The service provider has the capability of over ridding the power standby mode and continuing to support specific subscribers, or specific services such as video on demand.
One embodiment of the present invention supports power conservation functionality in two distinct methods. Firstly through use of statistical multiplexing. Any resources not actively used throughout the general architecture are automatically placed in a power standby mode. In this manner power consumption is minimized.
One embodiment is further able to conserve power during a power outage by reprovisioning the Voice over IP services. VoIP services are actually data services and require power to data portions of the server architecture. However, in case of a power outage at the subscriber premises, no other data services can be utilized. Thus by reprovisioning the data portion of the service provider server, the system is able to minimize its power consumption. Basic Voice over IP services will be maintained by use of power management control.
Another embodiment of the present invention further utilizes similar power managed re-provisioned voice over IP services at the customer premise side. Specifically, incorporated within one embodiment of the present invention is the capability of managed power for the voice over IP portion of the subscriber access equipment. The voice over IP portion of the subscriber access equipment is responsible for not only the packetization and compression, but is also responsible for the basic digitization and line interface to the telephony peripherals like a telephone or answering machine.
If the subscriber premises are experiencing a power outage condition then the subscriber can chose to maintain his basic data services by use of a universal power supply. The universal power supply provides the capability of supplying the power to the devices while no line power is available. An embodiment of the present invention requires the use of a universal power supply if the subscriber chooses to maintain data services. If this embodiment of the present invention detects a line loss condition then the system is placed into a standby mode. In the standby mode data services are disabled and the basic modem capability is disabled. The service provider server is notified that the subscriber access equipment is in a standby mode. After this point, when a telephony peripheral is detected to be off hook or an incoming call thru the service provider server is detected, the system enables the voice over IP portion of the subscriber access equipment and begins minimal modem functionality for the normal operation of the telephony peripheral. Battery backup circuitry within the subscriber access equipment is designed to be automatically charged during the times when power is available to the system. During the power outage condition, the system uses the battery backup portion to derive the necessary power.
One embodiment of the present invention incorporates an indicator which reports the state of the battery backup in a power outage condition when any telephony device is detected to be off hook. Furthermore, the system further has the capability to utilize standard flashlight batteries such as AA, C, or D size, to derive the necessary power during an power outage condition. In this manner, the system makes every possible attempt to maintain voice services even during an extended subscriber power outage condition.
An embodiment of the present invention incorporates the capability of caching internet based data at the service provider server. Web caching is used to improve the realized performance for the subscriber. The system maintains a custom per subscriber based table summarizing subscriber internet data usage. For frequently used data, or custom configured data usage, or push based internet data, the system request the data over the internet and holds it in a temporary buffer at the service provider server. Subsequently, when the subscriber accesses the data the service provider server cached data is delivered to the subscriber. In this manner the subscriber realizes the highest performance across the network.
One embodiment of the present invention periodically refreshes its cache content, based on network conditions or manual request by the subscriber. An internet wide refresh token scheme is utilized by the service provider server. When any internet data is cached, a refresh token is issued by the conforming hosting server. Subsequently if a server changes its page content, the previously requesting service provider servers are notified by use of issuing a new refresh coupon. The service provider server then schedules the need for refreshing the specific internet data. At subsequent favorable network conditions, the cache is refreshed with new internet data. In case, of a non-conforming hosting server, the service provider can configure the system of either self generate the refresh request in a variety for manners, such as once only, on a specific timing interval, or on a strictly network condition basis.
The advantage of the web caching to the subscriber is that the internet data is fetched from the hosting server during previous to the subscribers request. Upon receipt of the subscriber request, the service provider server is able to immediately forward the requested data limited only by the wire limitations of the subscriber line. In comparison, if the internet data had to be fetch across the internet, then the subscriber experiences the network delay of the internet, the hosting server delay, as well as the wire limitations of the subscriber line.
One embodiment of the present invention also incorporates a statistic reporting mechanism whereby the hosting server is notified of the caching results. The view statistics are important to the hosting server. When the service provider server request for a refresh coupon based caching session with a hosting server, the previous caching results and statistics are also transmitted. The service provider can configure the nature of the reported statistics such as view hits, among others.
An embodiment of the present invention incorporates the capability of a distributed server model. The prior art uses a central location for holding and storing subscriber data information. The present invention distributes intelligence across a host of service provider servers. The service provider servers are capable of subscriber line interfacing. Additionally, the service provider servers are able to respond to the requirements of the data, such as routing the data on a applications basis, or local storage, or local fetch requests.
The distributed server model also plays an intimate role in storage location decisions. Service provider administrative staff or subscribers can configure the service provider server to determine the storage usage based on an application basis. For example, e-mail data could be pushed down to the service provider server or subscriber premises; Video on Demand movie data could stay as part of the centralized network, and remote video recorder data could reside at the service provider server. One embodiment of the present invention incorporates a distributed storage model for the subscriber data.
The distributed server model of the present invention further enables a number of capabilities which otherwise do not exist in a centralized server model, such as in-expensive high performance web site hosting, and in-expensive high performance interactive voice response services. The use a plurality of servers to provide a given functional capability in the present invention enables it to provide a very high service capability. The distributed server model enables a high performance multi-service on a network level by virtue of the fact that a number of physically distinct elements are used to provide the functionality. Each of these elements has specific physical limitations, such as computational bandwidth and storage bandwidth. The distributed nature of the present invention enables it to utilize this capability in parallel. Whereas the centralized prior art approach is limited by the mandated serial nature of having to schedule the services in case of the limiting resources, such as storage access bandwidth, or computational bandwidth.
One embodiment of the present invention incorporates the capability of delivering content to the subscriber on a couponing basis. One embodiment architecture incorporated a smart card based debit capability, as well as a conditional access module capability. The smart card based debit capability incorporates a card reader, capable of verifying physical presence of either a smart card, a debit card, or and credit card, and reading its contents. In the case of a smart card, the debit is made directly from the card itself. Whereas in the case of a credit card, or debit card, the information is processed locally, however, the actual debit is made at the card holding banking facility. The conditional access module is responsible to decryption of the delivered content at the subscriber site. In one embodiment of the present invention the content provider provides encrypted data to the network by use of either terrestrial or satellite based transport.
Only subscribers authorized to use the data are forwarded the decryption coupon. The decryption capability coupled with the couponing mechanism is used for unidirectional secure transmission such as a broadcast mode or video on demand mode. In this configuration the data resides in an encrypted form within the service provider server and the content provider is guaranteed that the service provider does not have the ability to provide copies of this content to multiple subscribers and thereby circumvent royalties or other licensing arrangements. A bidirectional secure channel can be achieved by use of decryption/encryption conditional access module within the same architecture.
The smart card based debit card is also used to enable secure e-commerce. Within the scope of this document, e-commerce is used as a term relating to any form of a transaction between two parties or more that is not a physical transaction. Wherein, a physical transaction is used as a term relating to any form of a transaction between two parties or more that requires the physical presence of all parties. E-commerce does not require any transactions of monies, electronic or otherwise, but simply relates to any transaction where two parties or more wish to transact. One of the limitations of internet based business commerce, consumer commerce is a fact that the consumer has to issue a credit card number over the internet. It is feasible for an intelligent electronic criminal to steal the credit card number and make multiple charges against the credit card. The use of a debit card within one embodiment of the present invention requires a multi-step process for a e-commerce transaction to take place. Firstly, the authorization number needs to be disclosed by the subscriber requesting the transaction. Secondly, the physical presence of the smart card based debit card is required. Thirdly, the smartcard must have the credit to complete the transaction. The debit card can be used to limit the potential monies lost by the subscriber in case of a stolen authorization code. Furthermore, the card can be physically removed when the subscriber is not actively engaged in a transaction to minimize risk of misuse. In addition, the subscriber is able to interrogate the smart card to establish the outstanding credit balance within the smartcard. The smart-card reader can be enhance to also accept a debit card and a credit card, thus a physical requirement of the credit card is mandated to complete the transaction. The rest of the transaction occurs over a secure encrypted means and the transaction is validated at the transaction termination site.
One embodiment of the present invention incorporates a smart-card based immediate crediting capability within the network whereby a content provider or bandwidth provider can automatically deduct an authorized amount from the smart card. The smart card based debit or credit card capability is used within one embodiment to purchase temporary connective bandwidth such as making a long distance phone calls or temporary network bandwidth or performance. In one embodiment of the present invention the subscriber household or business is given several individual smart-cards. Each of these can be assigned to various individuals at the subscriber premises and in this manner individual usage and charge patterns can be monitored.
An embodiment of the present invention incorporates the capability of a statistically multiplexed modulation, demodulation, and channel processing. The Asymmetric Digital Subscriber line technology (ADSL) modulation requires a Fast Fourier Transform(FFT). The processing of the FFT is very computationally intensive. One embodiment incorporates the capability that the FFT processing is done on a demand driven basis. Thus, only those terms of the FFT which are non-zero are demanded to be computed. If a given carrier within the Discrete Multi-Tone (DMT) standard of the ADSL standard does not convey any information, its associated FFT term is not computed. One embodiment of the invention utilizes a shared processing pool for computational needs. Thus, the saved computational load, resultant from the uncomputed FFT term, is available for use by other subscriber line interfaces.
One embodiment of the present invention intelligently manages the carrier bit packing so that the minimal number of carriers are used by a subscriber line to transmit the required bandwidth subject to the associated signal to noise ratio. An embodiment minimizes the number of carriers and packs the maximum number of bits to deliver the service and thereby reduces the statistically multiplexed FFT computational load.
One embodiment of the present invention incorporates the use of a variable precision computation of the fourier transform. When only a few bits are packed within a carrier, a larger constellation size results in comparison to when more bits are packed within a carrier wherein a smaller constellation size results. However, the number of bits at the Analog to Digital converter and the Digital to Analog converter are fixed within the architecture of this embodiment. When fewer bits are being packed within a carrier this is because the system has a higher presence of noise. In this case, the maintenance of high precision within the FFT and other portions of the embodiment is futile, since the limiting factor becomes the noise within the channel. One embodiment of the present invention incorporates a variable precision FFT, where in the precision of the FFT is proportional to the packing density within the carrier. The precision maintained is more than the resultant noise at the carrier within the system. The variable precision FFT computation module enhances performance by utilizing the less computationally intensive lower precision computations wherever possible, and only computing the higher precision FFT carrier terms with high signal to noise ratio.
An embodiment of the current invention incorporates the capability of managing the quality of service (QoS) delivered to the subscriber. Traffic management in deployed at the service provider server. Several ATM traffic types are supported, in particular ABR, CBR, VBR, and UBR. ABR is an ATM traffic type where the packet is transported by the ATM switch on a Available Bit Rate means, wherein if the ATM switch has available bandwidth it forwards the packet. CBR is an ATM traffic type where the packet is transported by the ATM switch on a Constant Bit Rate means, wherein the ATM switch will forward the packet at a constant rate. VBR is an ATM traffic type where the packet is transported by a Variable Bit Rate means, UBR is an ATM traffic type where the packet is transported by a Unspecified Bit Rate means, wherein the ATM switch forwards traffic in an unspecified manner. One embodiment differentiates between ABR and UBR in that in ABR policy, the system gives a minimal bit rate guarantee with more bit rate capability if the network condition allow. In UBR policy, the system gives no minimal bit rate guarantee and data is forwarded with only a guarantee that it will be forwarded in sequential order within the network.
One embodiment of the present invention incorporates the capability of supporting multiple traffic types from the subscriber on an application basis. Furthermore, the subscriber is dynamically able to modify his bandwidth requirements so that he is able to achieve the his performance objectives without the need of the service provider intervention. An embodiment is able to deliver a bandwidth on demand capability by the use of dynamic provisioning in conjunction with the Quality of Service traffic management.
One embodiment of the present invention incorporates the capability of Remote Video Recording. The service provider server has the capacity to store selected recorded video content upon request by the subscriber. The present invention is differentiated from the prior art in that the prior art requires the subscriber to have an intelligent device at the subscriber premises to have the storage capacity to store the video content, whereas the present invention does not require such a intelligent storage capable device.
One embodiment utilizes the storage element within the service provider server to store the content of the video recording and then deliver this content to the subscriber in a demand basis. The Remote Video Recorder gives the subscriber capability of fast forwarding, rewinding, playing, pausing and stopping.
An embodiment incorporates within it the capability of video on demand which is functionally similar to a remote video recorder functionality, however, the main difference between video on demand and remote video recorder is that Video on demand data can be transmitted to any subscribe, whereas the remote video recorder data is only of interest to the subscriber. In one embodiment, a simple access device is used to interface to the service provider server, decode the video stream from the remote video recorder, and to interface to the television.
One embodiment of the present invention incorporates the ability for voice messaging services as well as a interactive voice response service. The algorithmic IVR capability and the voice messaging service capability reside at the service provider server. In one embodiment the subscriber has the capability to record custom greetings and options at the service provider server and be able to access these in a native telephone keypad based manner or a browser based interface.
An embodiment of the present invention incorporates a video conferencing capability. Prior art approach to video conferencing is typically done by use of ISDN lines or bonded multiple ISDN lines are brought into the subscriber premises. Dedicated hardware and software exist which then enables the user to be able to communicate with another video conferencing stations elsewhere. Prior art methodologies achieve full protocol compliant compression at the subscriber premises. The prior art approach is wasteful in that the subscriber based dedicated hardware can not be shared due to its physical location at the subscriber premises.
Another embodiment of the present invention segments the video conferencing capability into two pieces, the subscriber premise resident front end and the service provider server resident back end. The video conferencing front end is responsible for the Analog to Digital conversion and the Digital to Analog conversion as well as a cursory compression and formatting of the data to utilize the subscriber line. The video conferencing back end is responsible for full protocol compliance with the industry wide video conferencing standards such as H.320. The video conferencing back end is capable of interfacing to the video conferencing front end and is additionally responsible for the full protocol compression/decompression capability.
Another embodiment of the present invention has the capability to deploy desktop video conferencing within a business environment or a remote SOHO subscriber premises. SOHO subscriber premises capability can either be achieved by a dedicated stand alone back-end add in card and then subsequent use of typical DSL or ISDN interface, or by use of a service provider deploying the infrastructure of the present invention. Businesses can deploy video conferencing within the company premises by utilizing the standard PBX phone wiring and connecting it to the service provider server. In effect the standard PBX phone wire is then able to deploy DSL in the same manner as the service provider. In this manner, the businesses can easily deploy desktop to desktop video conferencing to every office within the premises.